Our overall objective is to continue to develop and implement high performance computer integration of pre-operative MR and CT data with intra-operative images in order to guide surgery intra-operatively. We will consolidate existing high performance computational hardware, networking facilities, segmentation and registration software, and develop user friendly interactive image processing software, for a system which can provide online, easily utilized information to surgeons during the execution of operative procedures. In order to accomplish these goals we will build enhancements to our existing systems for operator controlled segmentation (OCS), signal intensity-based segmentation (SIS) and template-driven segmentation (TDS). In addition, we will extend mutual information based (MI) rigid registration in various directions, with specific developments for use by applications in interventional MRI. We will also develop capabilities for non-rigid shape modeling and develop aggressive implementations of our basic non-rigid warping system. In order to test our image processing tools, we will build and utilize a validation database of cases.